View
14
Download
0
Category
Preview:
Citation preview
NGI Strategic Research Project SP2 2011-2013 Summary report, 19 November 2013 NGI report no. 20110001-R-01
Offshore wind turbine foundations
Main office:PO Box 3930 Ullevaal StadionNO-0806 Oslo, Norway
Street address: Sognsveien 72, NO-0855 OsloT: (+47) 22 02 30 00, F: (+47) 22 23 04 48 ngi@ngi.no
www.ngi.no
NGI (Norwegian Geotechnical Institute) is a leading international centre for research and consulting in the geosciences.
NGI develops optimum solutions for society, and offers expertise on the behaviour of soil, rock and snow and their interaction with the natural and built environment.
NGI works within the oil, gas and energy, building and construction, transportation, natural hazards and environment sectors.
NGI is a private foundation with office and laboratory in Oslo, branch office in Trondheim and daughter company in Houston, Texas, USA. NGI was awarded Centre of Excellence status in 2002.
32
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
Table of contentsBackground
OrganisationThe research programme was coordinated by NGI under the supervision of Vaughan Meyer. Technical guidance was provided by Knut H. Andersen (Technical Expert, Offshore Energy), Lars Andresen (Managing Director), Hans Petter Jostad (Technical Expert, Numerical Modelling), Karl Henrik Mokkelbost (Director, Offshore Energy), Per Magnus Sparrevik (Technical Expert, Subsea Technology) and James Strout (Head of Technology Development and Innovation).
The research team comprised:• Per Magne Aas• Lars Andresen• Wystan Carswell (Ph.D student –
University of Massachusetts Amherst)• Gustav Grimstad• Jörgen Johansson• Hans Petter Jostad• Amir Kaynia• Thomas Langford• Tom Lunne• Finn Løvholt• Christian Madshus• Karin Norén-Cosgriff• Ana Page• Joonsang Park• Morten Saue• Knut Schjetne• Per Magnus Sparrevik• James Strout• Hendrik Sturm• Maarten Vanneste• Inge Viken
An international external advisory panel provided industry input into the research programme and guidance of the research aims during the course of the programme. The advisory panel’s contribution to the research programme is greatly appreciated.
The panel members were:• Morten Liingaard (DONG Energy) - Denmark• Pieter Mattelaer (GeoSea) - Belgium• Henrik Mikkelsen (MT Højgaard) - Denmark• Julian Osborne (East Anglia Offshore Wind /
Vatenfall) – United Kingdom• Knut Ronold (DnV) - Norway• Marc Seidel (RePower) - Germany• Pål Johannes Strøm (Statoil) - Norway• Tor Inge Tjelta (Forewind / Statoil) - Norway
Table of contents
Background & organisation 2
Table of contents 3
Introduction 4 - 6
Quantitative shallow sub-surface characterization using seismics 7 - 9
Quantification of cyclic / dynamic load histories – cycle counting 10 - 11
Nonlinear Stiffness and Damping of Soils and Foundations 12 - 13
UDCAM and PDCAM: Soil models accounting for cyclic degradation 14 - 15
Instrumentation and Monitoring Strategy 16 - 18
Bibliography 19
The Norwegian Geotechnical Institute (NGI) initiated a
three-year (2011-2013) research programme with focus on
geotechnical issues specific to the offshore wind energy
sector. This research programme concentrated on ad-
vancing the current state of knowledge in the areas of site
characterisation, geotechnical design and monitoring of
offshore wind turbine foundations. This report presents the
main results of the research program. A bibliography of the
published papers is given in the last section of the report.
Those who are interested in more detailed information are
referred to these documents.
The Research Council of Norway and NGI provided fund-
ing for the research programme (total NOK 5.25 million,
spread equally over the 3 years).
Background
Contact: Vaughan Meyer , vaughan.meyer@ngi.no
54
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
IntroductionIntroduction
NGI involvement with offshore wind projectsWith the greater emphasis placed on the development of renewable energy in recent years, NGI has experienced a significant increase in work for offshore wind farm projects.
The nature of NGI’s offshore wind activities can be roughly equally split into areas of site characterisation, geotechnical design and monitoring.
Offshore wind farms and the turbine structures themselves present particular geotechnical challenges, which include:
• site characterisation of large offshore areas (up to approximately 8 500 km2) and associated spatial variability in geotechnical conditions
• the loading scenarios for geotechnical design of the foun-dations differ to those for other types of offshore structures
• foundation stiffness is important since the turbines are sensitive to motions as well as their frequency
• the level of industry experience with different foun-dation types varies
• the long-term behaviour of the different foundations is largely unproven
• there are uncertainties in the application of standard design approaches to some foundation types
• the foundations are a significant part of the overall structure cost, raising the prominence of the founda-tion design in the project
• given the large number of foundations to be in-stalled, robust designs and efficient installation pro-cedures are required
• environmental and code verification requirements may influence the chosen foundation design
Outline of researchThe focus of the research programme was to:• identify areas where geotechnical-related research
and development would be of benefit for offshore wind turbine foundations
• based on existing knowledge, improve existing meth-odologies and/or develop new techniques, focusing on the areas of site characterisation, geotechnical design and monitoring
• where possible, adopt an integrated approach to the research in conjunction with industry.
Research methodologyNGI’s approach to this research programme is based on over 40 years’ experience with foundations for offshore structures. In this respect, it is important to note that the change in focus from the design and construction of a foundation for a single, large offshore structure, such as an oil and gas platform, to the foundations for an offshore wind farm, is much like the shift from bespoke, or tailor-made engineering, towards mass production.
The research methodology has concentrated on fill-ing gaps in the existing knowledge base identified from the particular geotechnical challenges highlighted for offshore wind turbine foundations.
Introduction
Nature of NGI offshore wind farm project work (2008 - 2011)
Site Characterisation Monitoring
Geotechnical Design
O�shore Wind Turbine Foundations
• Philosophy• Geophysical techniques• Field (intrusive) investigations• Laboratory testing techniques• Reporting requirements
• Installation (operations and Metrology)• Operation (Long term monitoring)
• Load scenarios• Foundation types• Soil models• Calculation procedures• Soil-structure interaction
Geotechnical Design
33%
Monitoring39%
Site Characterisation
28%
Site Characterisation Monitoring
Geotechnical Design
O�shore Wind Turbine Foundations
• Philosophy• Geophysical techniques• Field (intrusive) investigations• Laboratory testing techniques• Reporting requirements
• Installation (operations and Metrology)• Operation (Long term monitoring)
• Load scenarios• Foundation types• Soil models• Calculation procedures• Soil-structure interaction
Geotechnical Design
33%
Monitoring39%
Site Characterisation
28%
Troll platform, Statoil Alpha Ventus (www.alpha-ventus.de Photo: Mattias Ibeler)
• Single structure (platform)• Cyclic loads from waves dominate• Big and robust structure• Large return on investment
• Multiple structures (OWTs)• Cyclic loads from wind and wave• Slender structure (serviceability issues)• Lesser return on investment
TAILOR MADE MASS PRODUCTION(alters basis for site characterisation, design, fabrication & installation)
Comparison between an offshore platform and an offshore wind farm
NGI engagements on offshore wind farm projects in Europe (2008-2011)
Research areas with subtopics
76
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
RationaleShear wave velocity, dynamic stiffness and soil damping are essential parameters for design of offshore windfarms. Traditionally, they are determined through laboratory tests, which are influenced by sample disturbance. Mapping lateral sub-surface variations over larger areas requires integration of non-invasive seismic data with geotechni-cal boreholes. Complementing conventional seismics with surface waves can be beneficial for geotechnical and engineering applications, particularly when cyclic load-ing or vibrations play a role. Surface waves are dispersive, which allows establishing shear wave velocity profiles down to about one wavelength. At given frequencies, multiple propagation modes exist. Using these in inversion improves resolution, reduces uncertainty, and extends the investigation depth. Multi-modal inversion relies on novel forward modelling that accounts for the overlying water. In addition, the quality factors, being inversely proportion-al to material damping, can be determined from near-surface geophysics and can assist in soil characterization.
Main Objectives• Evaluate marine surface wave data acquisition;
• Develop algorithms to obtain shear wave velocities from multi-modal surface waves;
• Estimate quality factors from near-surface seismics;
• Bridge the gap between geophysical and geotechni-cal parameters.
Survey design for high-quality surface wavesUsing surface waves offshore requires specific conditions on survey design (source and receiver configurations). A feasibility study, combined with a unique field experiment at Gjøa (Socco et al., 2011; Vanneste et al., 2011), reveals that the optimum acquisition consists of seabed-coupled shear wave vibrators with densely-populated (e.g., 2.5 m spacing) seabed-coupled ocean-bottom cables (Figure 1). Data at long offsets increase the possibility of detecting higher modes. A seismic source emitting low frequencies over which surface waves stand out yields higher-quality results (Figure 2). Alternative to the seabed-coupled system, a towed streamer (hydrophones or 2C) is feasible, but one quickly loses resolution and data quality drops with towing height above the seabed (Figure 3).
Forward modelling vs. inversion of multi-modal surface wavesInversion relies on proper forward models. Conventional forward models ignore the effect of the water column which results in erroneous shear wave velocities (Figure 4). Multi-modal inversion schemes that take this into account were not available when this R&D project started. Invert-ing multiple modes implies that the process becomes more data driven, and will yield improved results with lower uncertainties. We addressed these shortcomings by rewriting our in-house full-wavefield forward model Laysac to extract the kinematic solutions for multi-modal surface waves (Rayleigh, Scholte, Love), and implemented it into our inversion scheme (Figure 5).
Quantitative shallow sub-surface characterization using seismics
Researh projects
Title
Shallow Marine Surface Waves for Geotechnical Site Characterisation
Developing a NGI procedure for analysis of cyclic load - time data
Damping properties of soils
Dynamic soil-structure interaction
Integrated simulation of offshore wind turbines
Settlement of bucket and monopile foundations in clay
Cyclic ADP model
Attenuation from near-surface seismics
Analysis of soil damping and stiffness for Offshore Wind Turbines
Load assessment, geotechnical design and serviceability of a skirted OWT foundation on sand
Instrumentation and monitoring strategy for wind energy foundations
MASW: Joint multi-modal inversion, attenuation and 2-3D effects
Participation in the German FoU project: Offshore SI for GBS windmills
A procedure for establishing soil springs for structural design
Quantification of cyclic / dynamic load histories – Cycle Count
Global damping and stiffness response from OWT foundations
Upgrading and improvement of PDCAM and UDCAM
Year
2011
2012
2013
Research Area
Site characterisation
Geotechnical design
Site characterisation
Geotechnical design
Monitoring
Site characterisation
Geotechnical design
Project leader
Maarten Vanneste
Karin Norén-Cosgriff
Finn Løvholt
Jörgen Johansson
Hendrik Sturm
Morten Saue
Gustav Grimstad
Maarten Vanneste
Finn Løvholt
Hendrik Sturm
James Strout
Maarten Vanneste
Tom Lunne
Hendrik Sturm
Karin Norén-Cosgriff
Finn Løvholt
Hans Petter Jostad
Research ProgrammeThe research programme comprised a total of 17 individual research projects. Many of these projects represent a continuation of a
particular research topic through the three-year programme. Results from most of these projects are summarised in this report.
Survey design - Data examples
Figure 1. Gjøa field survey design for multi-component shear wave profiling (see Vanneste et al., 2011).
Figure 2. Gjøa shear wave data. (Top) Stacked shot gather. (Bottom) Spectral images converted into
phase velocity vs. frequency or wavelength. High-amplitude events are surface wave modes.
Marine surface waves
98
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
Application 1: High-resolution shear wave velocities and anisotropy The Gjøa field (North Sea, 360 m water depth) was the scene of a unique shear wave experiment (Statoil and NGI, 2007). The resulting multi-component seismic data, with 240 4C receivers at 2.5 m spacing (Figure 2, top) remain the only marine shear wave data available worldwide. Spectral images (Figure 2, bottom) reveal the fundamental and 4 higher surface wave modes. This is the only data set – both offshore and onshore – showing high-quality multi-modal surface waves, which makes it an ideal test case for forward modelling and inversion.
The data acquisition with different source and receiver orientations (in-line, cross-line) allows recording both Scholte and Love waves, which have particle motion in the vertical and horizontal plane, respectively. With the newly implemented inversion algorithm, we obtain larger penetration depths, and smaller uncertainties on the best-fit model. Combining the results by constraining the layer thickness in the inversion provides information on aniso-tropy, a feature typically ignored in site investigations. Our results indicate that shear wave velocity anisotropy can be up to 15%.
Application 2: shear wave attenuationSeismic wavefield propagation also depends on damp-ing, expressed as the inverse property, the quality factor Q. Soil damping is a driver for fatigue, particularly in monopile structures, but is difficult to determine. Soil damping consists of large-strain non-linearity and small-strain attenuation. The latter part can be determined by seismic data and can therefore be used to improve the prediction of soil damping. There is, generally, little known on attenuation values in shallow soils from geophysical measurements.
One approach, based on the half-power bandwidth method in frequency-wavenumber domain, measures the attenuation coefficient from the width of the spectral peak. The method can be applied to both the fundamen-tal and higher-order modes. The widths of the spectral peaks are largely controlled by the highest damping in the layered structure and less by its vertical variations.
From comparison of the spectral images of the Gjøa data with tuned full-waveform forward modelling, S-wave damping is less than 1% for the soft deepwater soils. The attenuation coefficient for the fundamental mode in-creases linearly as a function of frequency, indicating that damping is largely viscous rather than due to hysteresis (Figure 7).
Dissemination of results / Further readingFurther reading is given in the bibliography. NGI also hosted a workshop on attenuation in 2012, with partici-pants from the University of Southampton, the Norwegian Geological Survey, the University of Oslo, Duke University, and NORSAR.
AcknowledgementsThe Gjøa field data were collected in collaboration with Statoil. We also acknowledge the SEABED project C-Dog for additional data and testing.
Marine surface wavesMarine surface waves
Figure 5. Flow charts for surface wave acquisition, processing and inversion in order to obtain high-resolution shear wave velocity with depth.
Figure 6. (Left) Results from multi-modal Scholte (blue) and Love (orange) wave inversion, with constrained layer thickness. (Right) Ratio of horizontal and vertical velocities, as an indication of anisotropy (up to 15%) in the shallow sub-surface.
Figure 7. (Top) Cross-section of spectral image to deter-mine attenuation from surface waves. (Bottom) Attenu-ation coefficients for fundamental Scholte waves from Gjøa field data (blue) and tuned synthetic data (red).
Figure 3. (left): Normalised fundamental mode shape in the water column for a simple soil model and various water depths. This figure illustrates that both the receiver and source should – ideally – be coupled to the seabed to record surface waves over larger frequency range. Shear wave velocity is 0 m/s in the water and 100 m/s in the top soil (10 m). In the half-space underneath, shear wave velocity is 200 m/s.
Figure 4. Dispersion curve for fundamental Rayleigh (pink) vs. Scholte wave mode (other colours), illustrating the effect of the water layer (depth given in figure legend), for the same soil model as above.
Maarten Vanneste maarten.vanneste@ngi.no
Joonsang Park joonsang.park@ngi.no
Christian Madshus christian.madshus@ngi.no
Inge Viken inge.viken@ngi.no
Note: Surface wave forward models and inversion typically assumes a 1D layered soil structure, whereas the subsurface is essentially 3D. This assumption is acceptable as long as sub-surface structural variations remain limited over the spatial scale of the measurements.
Marine surface wavesMarine surface waves
1110
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
New methodA new method has been developed that can transfer general, often highly irregular wave and wind load data, to standard input format for soil cyclic degradation cal-culations. This transformation is to a single frequency load with separation of average and cyclic components and counting & grouping of loads into parcels of increasing amplitude and mean value.
Rain flow counting is often used in fatigue analysis of structures. In the rain flow method, all peaks are identi-fied and counted. However, the method allows ampli-tudes to be determined from local maxima and minima belonging to different cycles, which may lead to overes-timation of amplitudes, see Figure 2. As soil experiences kinematic and not isotropic hardening, the rain flow method is probably not so well suited for soil analysis.
In the new method, the amplitude of each half cycle is determined from adjacent local maxima and minima. The drawback is that the method requires operator input and is therefore more sensitive to operator judge-ment. One example is shown in Figure 3. The signal can either be counted as one cycle with a high amplitude disregarding the smaller internal cycles, or it could be counted as many cycles with lower amplitudes. The
new method gives the operator the ability to override the maxima and minima identified by the program and define cycles interactively on the screen.
In order to evaluate the effect of different ways to define cycles on soil shear strain accumulation, the method uses NGI’s in-house program ACCUMUL to calculate resulting equivalent number of cycles for different cycle definitions.
The new method has been tested on a simulated load time series and the result compared with that obtained from a rain flow analysis on the same load time series. The comparison shows that the deviations between the two methods are considerable, in terms of both the iden-tified number of cycles in the time series and the calcu-lated equivalent number of cycles. Correct definition of load cycles is very important for design of the structure and foundation.
Figure 1. Converting load time series to standard input format for soil cyclic degradation calculations
Quantification of cyclic/dynamic load histories – cycle counting
Cyclic/dynamicCyclic/dynamic
Karin Norén-Cosgriff karin.noren-cosgriff@ngi.noFigure 2. Rain flow method versus NGI’s approach Figure 3. Different ways to define cycles affects the
determined amplitudes
Figure 4. Comparison between the new method (above) and the rain flow method (below).
1312
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
The periodic forces from wind, surface waves and swells acting on offshore installations, such as OWTs, lead to cy-clic motion of the foundation soils. The damping charac-teristics of the soil results in energy dissipation, which in turn alters the cyclic response of the structure. Understanding damping is therefore important for the cost-effective de-sign of structures. OWT foundation damping has not been widely studied, and consequently the influence of soil damping on OWT structural response is poorly understood. Research undertaken by NGI provides a step forward towards a better understanding and improved interpre-tation of stiffness and damping parameters for soils and foundations.
Intrinsic soil damping under combined permanent and cyclic loadOWT foundations are typically subject to combinations of cyclic and average loads from wind and waves. In addition, the different loads may have different frequen-cies and approach from different directions. Hysteretic soil damping enters only through the cyclic part of the stress-strain curve during a load cycle, but often an aver-age load is also present, leading to strain accumulation. Figure 1 shows a typical cyclic stress-strain curve, as well as an example laboratory measurement including both cyclic and average stresses. This strategic R&D program has shown that the measured hysteretic soil damping is influenced by the accumulated strain, and that this must be properly accounted for to interpret the soil damping correctly in OWT foundation design.
Nonlinear Stiffness and Damping of Soils and Foundations
Figure 1. Left panel, simplified sketch of the relation between cyclic stress and cyclic strain for pure cyclic loading. The damping factor D is based on the area of the stress strain curve (external work) over the total potential energy. Right panel, real laboratory measurement data combining permanent and cyclic loads.
Figure compilation:
A new method for deriving soil damping from the laboratoryA new method for quantifying soil damping from labora-tory tests has been developed. Analyses of laboratory data have shown that the interpreted hysteretic soil damping has historically been overestimated for large accumulated strains. This is visualized in Figure 2, where damping as a function of cyclic strain level are compared for the new and old interpretation methods. The overesti-mation occurs when the stress strain loops are open due to strain accumulation, as illustrated for the clay sample. Figure 2 further demonstrates the damping behaviour for different soil types e.g. sand and clay. Whereas damping increases with the maximum cyclic strain for both clay and sand up to a certain strain level, an opposite trend is found at larger strains for sand. The shape of the stress-strain loop largely determines this behaviour.
Effect of the foundation nonlinearity on monopile Offshore Wind Turbines OWTs are lightly damped structures, often with fatigue governing the design. Therefore, a thorough understand-ing of different damping sources, such as aerodynamic, hydrodynamic, structural, and foundation is essential for a cost effective design. Foundation damping is recently attaining more interest from the OWT industry. Foundation stiffness and damping depends not only on the soil prop-erties but also foundation geometry, load intensity and frequency. As the damping is load dependent, the prob-lem becomes non-linear and must be evaluated using realistic soil models. Examples of simulated cyclic motion of a monopile foundation, incorporating damping, are shown in Figures 3 and 4. The global damping of the entire OWT structure calculated from the simulations for a rotor stop scenario (see figure 5) is found to be comparable to those obtained from field measurements. Foundation damping gives an important contribution to the global system damping in addition to e.g. tower oscillation dampers. Radial spreading of stress waves in the soil does not contribute much to the global dynamic response of the first mode of OWTs. However, intrinsic soil damping as well as the effect of added mass do affect the overall response, which may be utlized in OWT design.
Soil DampingSoil Damping
Figure 5. Offshore wind turbine tower free vibration numerical simulation.
Figure 2. Upper panel, damping ratios and examples of cyclic curves for a clay sample. Lower panel, damping ratios and examples of cyclic curves for a sand sample. The insets show examples of single load cycles at a given strain level.
Figure 3. 3D model for evaluating dynamic response of foundation
Figure 4. Rotation of monopile foundation with nonlinear soil behavior
Finn Løvholt finn.lovholt@ngi.no
Jörgen Johansson jorgen.johansson@ngi.no
Christian Madshus christian.madshus@ngi.no
Wystan Carswell wcarswel@engin.umass.edu
wp=1/2⋅γc⋅τc
pwwD ∆
⋅==π
η41
2
∆w
τ
γγc
τc
wp
1514
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
UDCAM and PDCAM: Soil models accounting for cyclic degradation
ApplicationsUDCAM and PDCAM are applicable for general bound-ary value problems and have been proved especially suitable in the design of:
• Monopiles, where the cyclic degradation of the soil varies along the pile. This is important for wind tur-bines structures, since the permanent rotation of the structure can govern the design (figure 3).
• For jack-up structures on bucket foundations: it is important to assign the correct rotational soil stiffness to these structures, since the moment fixity at the bottom of the legs can govern structural utilisations in the legs (figure 4).
What is cyclic degradation?
Figure 2. Results of a model test with monotonic and cyclic loading on a gravity platform on clay. The results show that the stiffness and strength are lower for cyclic loading and that the displacements increase with the number of cycles.
Offshore structures are subjected to combined static and cyclic loading due to the weight of the structure, wind, current and waves. The effect of cyclic degradation in the soil during these load conditions may be significant and therefore needs to be properly taken in to consideration.
How do we account for cyclic degradation?We analyse the behavior of the soil under cyclic loading based on non-linear stress-strain relationships from cyclic contour diagrams. These contour diagrams are estab-lished from laboratory tests, and they contain information about the reduced strength, the increased cyclic strain amplitudes and increased permanent strains as function of number of cycles at different cyclic and average shear stress levels. Figure 1 shows an example of a cyclic contour diagram.
In order to calculate the cyclic degradation of the soil, the real cyclic load history is transformed into an idealised loading composition, where the design storm is divided into parcels of constant cyclic load amplitudes. An equivalent number of cycles, Neq, is used as a memory of the cyclic effect. A high Neq implies a high cyclic degradation of the soil, while Neq = 1 means no cyclic degradation. Figure 4 shows the calculated distribution of the equivalent number of cycles, Neq, at the end of an applied cyclic load history.
This procedure has been implemented into the finite ele-ment code PLAXIS as the UnDrained Cyclic Accumulation Model (UDCAM) and the Partially Drained Accumulation Model (PDCAM). These models allow us to apply the ad-vantages of the FEM in combination with our well-proven calculation procedures. We can then calculate the cyclic degradation of the soil in any point of the soil and model the capacity and displacements of any foundation sub-jected to cyclic loading.
UDCAM / PDCAMUDCAM / PDCAM
Figure 1. Example of a cyclic contour diagram. The strength, defined by the failure line, decreases significantly with the number of cycles.
Figure 4. Distribution of the equivalent number of cycles, Neq, at the end of an applied cyclic load history for a bucket foundation. Higher values of Neq imply higher cyclic degra-dation of the soil.
Hans Petter Jostad hans.petter.jostad@ngi.no
Ana Page ana.page@ngi.no
Hendrik Sturm hendrik.sturm@ngi.no
Cyclic degradation is the reduction in strength and stiffness
of the soil due to the generation of pore pressures under
undrained and partly drained conditions and destructuration
during cyclic loading. The generation of pore pressure results in
reduced effective stresses in the soil and the development of
permanent strains.
Figure 2 shows load-displacement curves of a model test with
monotonic and cyclic loading on a gravity platform. The figure
shows the reduced capacity during cyclic loading compared
to monotonic loading, and the increase of the cyclic displace-
ment amplitudes (or reduction in stiffness) and increased
permanent displacement of the foundation with the number of
applied load cycles.
Figure 3. Contour plot of total cyclic shear strains at the end of the applied load history for a monopile foundation
1716
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
Instrumentation and Monitoring Strategy
These foundation solutions have many similar monitor-ing/instrumentation needs, as well as specific monitoring needs for piled foundations and caisson foundations:
For all types of foundations:• Wind
• Waveheight
• Tiltoftower
• Scourandcurrents(iftheseabedisproneforsedi-ment transport)
• Cyclicporepressurealongfoundationelements(pile or caisson)
• Strain/fatigueincriticalstructuralmembers
• Dynamicmotionofthestructure/toweratvariouslevels
Driven pile foundations (monopiles or jackets with three or more piles):• Axialstrainalongthepile(P-Ybehaviour)
• Lateralearthpressurealongthepile(difficult)
• Internalcorrosiontransitionpiece-piletop
• Deformations/strain/integrityofgroutedconnectionsand transition pieces
Caisson foundations (monopods or tripod/quadropod jackets):• Straininconnectionsbetweentower/jacket
structural elements and the caisson
• Dynamicmotion(rotation/linear)ofthefoundation
• Loaddistribution,verticalearthpressurealongthebase of the caisson (if not grouted)
• Settlement(shakedown)
Technology application example: pore pressure measurement in caissonsPiezometers suitable for submerged installation generally consist of heavy duty filters at depth, communicating via piping to a measurement point positioned above the caisson (often on top of it). For skirted foundations it is usually the pore pressure at either sides of the skirt tip (0.5-1m from the tip) and the caisson pressure at the base which are of primary interest for monitoring (both direct response to transient overturning loads and pos-sible cyclic pore pressure accumulation).
The main objectives of this research component were as follows:
• Evaluate monitoring needs specific for the wind
energy industry, related to type of installation. Un-
derstanding the monitoring needs is the basis for the
monitoring strategy
• Review specific technologies and instruments rel-
evant for the monitoring needs to identify suitable
system solutions to implement the strategy
• Consolidate experience from a related industry (sub-
sea oil and gas installations) to develop guidelines
for the practical design, procurement and installa-
tion of appropriate monitoring systems.
Objectives
Background Offshore Wind Turbine Structures are tall and slender, thus relatively sensitive installations in regards to structural loading and structural response. In particular when water depths and turbines are getting bigger the dynamic motions, settlements and possible tilting are particularly onerous for the generator and rotors. In addition, con-firmation of foundation and structural performance are essential to verify design assumptions for large scale field developments. Finally, the instrumentation and monitor-ing data from current installations provides insight and a technical basis for optimizing future foundation and structural solutions.
Understanding monitoring needsPresently the alternatives for offshore wind turbine foun-dations can be divided into three broad categories:
1. Monopiles /monopods (single pile/single suction anchor)
2. Jackets or tripods, with piles or suction caissons
3. Gravity base structures with skirts
InstrumentationInstrumentation
Figure 1. Monopile foundation (left) and tripod jackets with pre-driven piles or suction caissons (right)
Figure 3. Position of piezometer filters on the caisson and routing of hydraulic lines to the sensor heads
Figure 2. Caisson piezometers (8) on a tripod jacket with caisson foundations
The hydraulic piping from the piezometer filters are routed to a sensor head containing differential pres-sure sensors and other electronics. The sensor head can be hooked up in advance or after driving (piles). The termination can be equipped with a solenoid operated bypass valve (opens the piezometer line to sea) allow-ing for de-airing of the line and zero point check of the
differential pressure sensor. By means of using differential pressure sensors and hydraulic lines saturated with sea-water, the pore pressure is directly recorded and com-pensated for tidal and atmospheric pressure variations. Multiple piezometers and other sensors such as acceler-ometers can be integrated in the sensor head optimizing the configuration of the monitoring system.
1918
Offshore wind turbine foundations, NGI Strategic Research Project SP2 2011-2013, Summary report.
Bibliography
Page Risueño, A. M., Jostad, H. P. & Saue M. (2013)Application of an undrained and a partially drained cyclic accumulation model for monopole design5iYGEC’13 - 5th International Young Geotechnical Engineering Conference, Paris, France
Sauvin, G., Bazin, S., Vanneste, M., Lecomte, I. & Pfaffhuber, A.A. (2011) Towards joint inversion/interpre-tation for landslide-prone areas in Norway - Integrating Geophysics and Geotechnique. Near Surface 2011, 17th European Meeting of Environmental and Engi-neering Geophysics. EAGE, Leicester, UK, pp. 5.
Socco, V.L., Maraschini, M., Boiero, D., Vanneste, M.,Madshus, C., Westerdahl, H., Duffaut, K. & Skomedal, E. (2011) On the use of NGI’s prototype seabed-cou-pled shear wave vibrator for shallow soil characteriza-tion - Part II: Joint inversion of multi-modal Love and Scholte Surface Waves. Geophysical Journal Interna-tional 185, pp. 237-252.
Sturm, H. (2011)Geotechnical performance of a novel gravity base type shallow foundation for offshore wind turbinesGeotechnik, Volume 34, Issue 2, pp. 85–96.
Sturm, H. (2012)Ausgewählte Aspekte zur Bemessung und Installation von Pfählen für Offshore Windenergieanlagen10. Tiroler Geotechnik- und Tunnelbautag
Sturm, H., Andersen, K.H., Langford, T. & Saue, M. (2012)An introduction to the NGI cyclic accumulation ap-proach in the foundation design of OWTs12th Baltic Sea Geotechnical Conference, Infrastruc-ture in the Baltic Sea Region Rostock, Germany 31 May - 2 June 2012
Vanneste, M., Longva, O., L’Heureux, J.-S., Vardy, M.E., Morgan, E., Forsberg, C.F., Kvalstad, T.J., Strout, J.M., Brendryen, J., Haflidason, H., Lecomte, I., Steiner, A., Kopf, A., Mörz, T. & Kreiter, S. (2013) Finneidfjord, a Field Laboratory for Integrated Submarine Slope Stability Assessement and Characterization of Landslide-prone Sediments: a Review. Offshore Technology Confer-ence (OTC2013). OTC, Houston, TX, pp. 17.
Vanneste, M., Madshus, C., Socco, L.V., Maraschini, M., Sparrevik, P.M., Westerdahl, H., Duffaut, K., Skomedal, E. & Bjørnarå, T.I. (2011)
On the use of NGI’s prototype seabed-coupled shear wave vibrator for shallow soil characterization - Part I: Acquisition and processing of surface wavesGeophysical Journal International 185, pp. 221-236.
BibliographyInstrumentation
Publications in scientific journals
NGI reports
Considerations for the selection of sensorsSome basic guidelines for sensor selection include:
• Parametertobemeasured:Determinethebestmethod to obtain the required parameter(s) including how to get the desired accuracy and resolution.
• Measuringrange,precisionandaccuracyrequired.Costis a function of the specifications - choose the specifica-tions appropriate for the design and overall monitoring performance and not simply the very best sensor on the market (the sensor may not be the limiting factor).
• Priority:Whichprioritydoyougivetothisparticularmeasurement? This may govern the type of equipment you choose with respect to price and redundancy.
• Duration:Forhowlongshallthemeasurementpro-gram last? Type of equipment, choice of materials, etc. will depend on this. Bear in mind, however, that a successful monitoring program which gives interesting data is often extended – be prepared for this.
• Environmental:Theenvironmentalconditionsmustbetaken into account when choosing materials, rugged-ness of enclosures, barrier philosophy, physical mount-ing points and similar mechanical design properties.
• Signaltype:Whichsignaltype(frequency,voltage,cur-rent, digital, optical etc.) is best suited for this particular application? Noise and cable lengths, will this system interface with other systems, and are there existing data architecture/interfaces which have to be complied with?
• Sensormaterials:Requirementsregardingcorrosion,pressure, size, electrical effects etc.
• Sensormanufacturer:Previousexperiencewithsupplier.
• Modificationsandspecialcalibrations:Aretheintend-ed sensors / instruments to be used under conditions outside of normal specifications? Contact the manu-facturer or implement a dedicated test/verification program to establish suitability of the technology.
Summary experience: The ‘10 commandments’ of subsea instrumentation
Dissemination of resultsThe results of this study are introduced as part of the ‘Best Practice’ for instrumentation and monitoring system design at NGI.
Our work is also made available to commercial proj-ects providing design advice and monitoring solutions for offshore wind energy development.
Per Sparrevik per.sparrevik@ngi.no
James Strout james.michael.strout@ngi.no
Figure 4. Left: Data record from Ormen Lange down hole piezometer slope stability assessment with the sensor head at the seabed. The differential pressure sensor was equipped with solenoid operated bypass valve for automatic zero point checks and de-airing. Right: Data example from Femern large scale tests pile piezometers showing pore pressure dissipation after driving.
NGI Report No. 20110087-00-1-RIntrinsic and global damping of soils and OWT foundationsRev. 0, In preparation
NGI Report No. 20110186-00-1-R Soil Structure Interaction for OWT monopile foundation Rev. 0, dated 7 December 2011
NGI Report No. 20110255-00-1-R A new procedure for analysis of load 20130098-01-R SP2.33 Quantification of cyclic / dynamic load histories - Cycle CountRev. 0, dated 19 December 2011 NGI Report No. 20120132-01-RSP2: Instrumentation and monitoring strategy for wind energy foundations - Instrumentation and monitoring strategyRev. 0, dated 18 January 2013
NGI Report No. 20130091-01-RAn improved procedure for establishing cyclic soil properties of clays and sandsRev. 0, In preparation
1. Provide for adequate planning and concept design development
2. Design for harsh conditions and rough handling
3. Plan for contingency, redundancy and back-up
4. Maintain barriers and control corrosion
5. Perform functional testing
6. Simplify the installation approach if possible
7. Work closely with the offshore contractor
8. Meet the delivery schedule
9. Make the data available
10. Mind the devil (he is in the details)
Recommended